Moving from Another Language to LISP?
Steve Graham
their application. It currently uses a B-tree database and involves
getting orders from another system, getting results from automated
instruments, and packaging those results off to the 1st system. Cost of
the language/database vendor is one concern and wanting to move from
this 3rd generation, albeit very productive, language into the more
modern world of object-oriented programming is another. Currently Java
seems to be a/the leading contender.
If you were trying to convince the management to move to LISP, what
would you say or do?
Thanks, Steve
Kenny Tilton
and to debunk the many myths -- i am guessing your friend knows all that
stuff -- but it would also help to take a few weeks of their own time to
whip up a rough prototype with as much core functionality as possible. I
had the latter in hand when I told the boss we should be using lisp.
sounds like your friend needs to price the tools as well. linux and
cmucl would be free, fwiw, but even commercial stuff is not too
expensive and management might get a warm fuzzy off a commercial vendor.
kenny
clinisys
cr88192
I think I will post to the same message on comp.lang.smalltak only varying
one word...
--
<cr88192[at]hotmail[dot]com>
<http://bgb1.hypermart.net/>
Vassil Nikolov
> I think I will post to the same message on comp.lang.smalltak only varying
> one word...
It had already been done. But not to any other group---is
Smalltalk the only competitor to Lisp when it comes to finding a
non-Java solution?
---Vassil.
--
Garbage collection is charged at 0.19e-9 cents a cons. Bulk rates
are also available: please contact memory management for details.
ilias
> Steve Graham wrote:
>
>
>>I have a programmer friend whose division is considering a rewrite of
>>their application. It currently uses a B-tree database and involves
>>getting orders from another system, getting results from automated
>>instruments, and packaging those results off to the 1st system. Cost of
>>the language/database vendor is one concern and wanting to move from
>>this 3rd generation, albeit very productive, language into the more
>>modern world of object-oriented programming is another. Currently Java
>>seems to be a/the leading contender.
>>
>>If you were trying to convince the management to move to LISP, what
>>would you say or do?
>>
>
> I think I will post to the same message on comp.lang.smalltak only varying
> one word...
>
you have to declare that.
you posting is missleading.
why do you do that?
Pascal Costanza
> If you were trying to convince the management to move to LISP, what
> would you say or do?
Recently, Richard Gabriel has given a short summary of arguments for
Common Lisp at http://groups.yahoo.com/group/feyerabend-project/message/252
Here are some highlights:
"My team will be able to program circles around everyone else. They
will be able to construct rapidly a language specific to the problem
we are solving rather than using a language designed by computer
scientists worrying about their place in history and a herd of
library writers working in cubicles a thousand miles from our
business."
"I will be able to point to various examples where Lisp programmers
have written not only 3-5 times faster, but they wrote things other
programmers thought were impossible."
"Because Lisp is dead, I'll get better programmers for less money.
I'll be able to guarantee 50 more IQ points for the same pay."
"I'll be compatible with everything because it is right now. And if
someone throws me a bug, I can code around it in a few minutes."
I agree with most of this, except for Lisp being dead. ;)
Pascal
cr88192
> On Sat, 05 Oct 2002 19:16:53 -0700, cr88192
> <cr8...@hotmail.nospam.com> said:
>
> > I think I will post to the same message on comp.lang.smalltak only
> > varying one word...
>
> It had already been done. But not to any other group---is
> Smalltalk the only competitor to Lisp when it comes to finding a
> non-Java solution?
>
the original poster had posted nearly identical messages to both groups, I
had thought if would be funny to do similar as a reply to each version...
not only that, but I would make each message refer to the one on the other
group.
cr88192
actually I typed both at roughly the same time, each being a reply to the
respective message.
if you notice the same variations occure in the quoted versions as in the
original messages, as such this was not a repost...
the one word was the usenet group, each pointing to the other.
I had figured it would have been funny to someone who reads both groups...
Vassil Nikolov
*> Vassil Nikolov wrote:
>> On Sat, 05 Oct 2002 19:16:53 -0700, cr88192
>> <cr8...@hotmail.nospam.com> said:
>>
>> > I think I will post to the same message on comp.lang.smalltak only
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> > varying one word...
>>
>> It had already been done. But not to any other group---is
>> Smalltalk the only competitor to Lisp when it comes to finding a
>> non-Java solution?
>>
*> the original poster had posted nearly identical messages to both groups, I
*> had thought if would be funny to do similar as a reply to each version...
I see now. I read the message to which I responded as if it said
`I will post the same message...' while in fact it meant `I will
^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^
respond to the same message...'
^^^^^^^^^^^^^^^^^^^^^^^^^^^
And my question (whether only Lisp and Smalltalk can be considered
for a non-Java solution) refers to the fact that the original posts
of those nearly identical messages were in just those two
newsgroups.
ilias
a joke is a joke.
modifying a posting of another person is not a joke.
Erik Naggum
| If you were trying to convince the management to move to LISP, what would
| you say or do?
I have no experience with convincing management to move to Common Lisp,
but as a consultant to companies who already had failure on their hands,
or a problem to which they did not know a solution, offering Common Lisp
and a significantly better chance to succeed has not been met with much
resistance. Now, as a consultant, they may have trusted /me/ rather than
my tools and may well have called me in the first place to help them out
of their predicament, so what experience I have may not be all that useful
to somebody else. However, one thing should be generally applicable:
Size up the problem and spend as little company resources as possible to
determine that you can solve the problem. This may involve solving part
of it for "free", just to demonstrate how you would do things. This
achieves two separate goals. The first is to show your management that
you know your stuff and that they should trust you with the problem. The
second is to show your management that you are willing to take risks of
your own in order to "prove" something to them. Normally, both of these
are valued by your management. If not, seek employment elsewhere, as
they do not trust you or appreciate your efforts to solve their problems.
To really succeed, tell your management about what you intend to do up
front, and show that you can do the preliminary work without involving a
lot of people and that you can deliver what you promised on time, then
show them what you need in order to finish the task. If you have a track
record of delivering on time, that also helps a lot.
Note that this stresses "do" over "say". "Say" alone never cuts it.
--
Erik Naggum, Oslo, Norway
Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.
Steve Graham
c.l.l and c.l.smalltalk. I have virtually no experience in OO, but I
have followed LISP and Smalltalk with interest, and thought readers of
the 2 newsgroups might have some valuable input to the discussion. I
chose not to cross-post the message: Who wants to have their newsgroup
cluttered up with the traffic from another.
Thanks to all those who have posted their suggestions.
Steve
===
ilias
*sorry*
my fault.
for some reason i don't see the original post of "Steve Graham" in this
newsgroups.
so i thought you copy his message, change the word (smalltalk => LISP)
an post a reply here.
again, sorry for that.
cr88192
I did not modify the other's message, there were originally 2 nearly
identical posts on 2 groups, and I served to write 2 nearly identical
replies.
cr88192
>> a joke is a joke.
>>
>> modifying a posting of another person is not a joke.
>>
>
> *sorry*
>
> my fault.
>
> for some reason i don't see the original post of "Steve Graham" in this
> newsgroups.
>
> so i thought you copy his message, change the word (smalltalk => LISP)
> an post a reply here.
>
> again, sorry for that.
accepted.
cr88192
> There has been a lot of talk about my having posted the same message to
> c.l.l and c.l.smalltalk. I have virtually no experience in OO, but I
> have followed LISP and Smalltalk with interest, and thought readers of
> the 2 newsgroups might have some valuable input to the discussion. I
> chose not to cross-post the message: Who wants to have their newsgroup
> cluttered up with the traffic from another.
>
> Thanks to all those who have posted their suggestions.
>
yes, and my attempt to use the fact as a base for a joke seemed to
backfire...
sorry I couldn't really add much useful to the conversation, I am off a bit
messing with my own prog lang, which I have not been able to put too much
work into recently as I have been quite busy otherwise...
Ng Pheng Siong
> I have no experience with convincing management to move to Common Lisp,
> but as a consultant to companies who already had failure on their hands,
> or a problem to which they did not know a solution, offering Common Lisp
> and a significantly better chance to succeed has not been met with much
> resistance. Now, as a consultant, they may have trusted /me/ rather than
> my tools and may well have called me in the first place to help them out
> of their predicament, so what experience I have may not be all that useful
> to somebody else.
I read this to mean some of these companies calling on you were not using
Common Lisp in the first place.
How do they (or you) deal with the Lisp code maintenance issue after you're
done?
Thanks. Cheers.
--
Ng Pheng Siong <ng...@netmemetic.com> * http://www.netmemetic.com
Christopher C. Stacy
Ng> According to Erik Naggum <er...@naggum.no>:
>> I have no experience with convincing management to move to Common Lisp,
>> but as a consultant to companies who already had failure on their hands,
>> or a problem to which they did not know a solution, offering Common Lisp
>> and a significantly better chance to succeed has not been met with much
>> resistance. Now, as a consultant, they may have trusted /me/ rather than
>> my tools and may well have called me in the first place to help them out
>> of their predicament, so what experience I have may not be all that useful
>> to somebody else.
Ng> I read this to mean some of these companies calling on you were not using
Ng> Common Lisp in the first place.
Ng> How do they (or you) deal with the Lisp code maintenance issue
Ng> after you're done?
I would assume that they deal with it in the same way as anything
else. They either hire the original programmer again, or if they
don't want to do that, they hire someone else who has the same skills?
So I suppose I don't understand the question.
Umm, is there some other way of doing it?
Erik Naggum
| How do they (or you) deal with the Lisp code maintenance issue after
| you're done?
With support contracts. I always ensure that I have people who can take
over for me. Sometimes, customers have wanted to rewrite things that
work into a more "mainstream" language. These projects have always
failed, however. I find that moderately amusing, because no matter how
hard you try to blame someone else for your failure, a working system
simply and effectively mocks every attempt. People whose purpose it is
get their job done, do not desire to rewrite systems because some new
manager has been to a Java conference and needs to defend the expenses.
Kaz Kylheku
What issue would that be, and how is it different from a Java, Perl,
Python, Visual Basic or C++ code maintenance issue?
There are lots of Common Lisp programmers who will gladly bail their
C++ or Java job to get paid to work with the intelligent language that
they know. If you hire these people, you are not only getting a Lisp
programmer, but a C++, Java or whatever programmer in the same
package. This individual will be grateful for an opportunity to write
software right, and work hard for you.
If you advertize a Lisp job anywhere, it will nicely stand out,
because it won't be listed alongside three hundred identical postings,
and whereas the stack of applications that you get will probably be
substantially smaller than if you advertised a (for instance) C++ or
Java development position, it will have a far lower proportion of
unqualified applicants.
As a last resort, you could always offer the job to someone who is
well qualified, having a track record in design and implementation
using some other languages, and who is willing to pick up Lisp. That
person will probably make some mistakes, such as choosing poor data
representations and approaches, but the result will likely still be
superior, and as that individual learns more, Lisp will allow the old
design decisions to be refactored with great ease.
Unpopular programming languages do not attract unqualified people in
droves. You get the odd moron and a bit of lunatic fringe (consider
recent Usenet examples such as of ``ilias'' and ``gnuist007'') but
that's it.
Remember, not everyone who claims to be a programmer and know some
languages is actually a qualified developer who is fit to maintain
your code, regardless of what programming language it is written in.
The future of that code depends on on having *something* done to it,
but having something *right* done to it.
Ng Pheng Siong
Putting it the way you do, it does sound like a silly question. ;-)
My question was in the context of a consultant coming into a company
without prior Lisp experience and concurrently facing failure in some
project, where consultant then delivers some end product in Lisp. Erik has
answered that very clearly. (Thanks.)